On 5/24/06 6:06 AM, "Sharon" <sharonanandhi at="" gmail.com=""> wrote:
> Hi,
>
> I am looking for a way to load background corrected data (all the 24
arrays
> in a single file) into limma. This is a 2-channel array data, so I
have 3
> columns (cy3, cy5 & flag) for each array with probe name & ID. I
have tried
> to load the data with read.table(). Now, do I need to convert this
table
> into RGList?
You need to make 3 separate matrices for cy3, cy5, and, if you want to
use
flag, a "weights" matrix; all of this can be done by subsetting out
the
columns that you want from your "big table". Then you can construct
your
RGList by hand. You will of course probably want to add a "targets"
data.frame as well as gene information. After that, everything should
proceed as normal.
Sean
On 5/24/06 9:39 AM, "Sharon" <sharonanandhi at="" gmail.com=""> wrote:
> Hi Sean,
Thanks for the suggestions. As you suggested, I have created 3
> different
matices (data.cy3, data.cy5, data.flag) from the large table. To
> make a
RGList, I looked at read.maimages() in limma. Though, I can make a
list
> with
3 matrices, I am still not sure about making 'targets' filed. It will
be
> of
great help if you send some info on making RGList by hand.
Thanks again
> for your help.
Following is the code, which I have used:
all <-
> read.table("data.txt", sep="\t", header=TRUE)
data.cy5 <- all[,
> c(seq(3,74,by=3)] # have 24 slides
data.cy3 <- all[,
> c(seq(4,74,by=3)]
data.flag <- all[, c(seq(5,74,by=3)]
genes <-
> all[,2]
Sharon,
You can do something like:
RG <- new("RGList",list(R=data.cy3,G=data.cy5,targets= ....))
You can include in your list all the things you are used to seeing
(weights,
targets, genes, etc.) as long as you are sure that the ordering of
columns/rows is appropriate. This may take some checking and/or data
manipulation to make sure that this is the case, but the call shown
above
will make you a full-fledged RGList object.
Sean
